how many DACS ? new setup

usb dacs any audiophile grade ?

Hi,

I need a PC audio setup for music listening, got some new bookshelf actives

Ive been away from audio fro over 5 yrs

What a shock when I saw USB based sound cards and DACS

I cant believe that USB is acceptable

prove me wrong, please..it will diminish the nightmares using USB as a connector

How is it done anyways ?
 
Last edited:

Billyo

Member
2011-04-15 1:12 pm
Sydney
I used a DAC connected to spdif for a while, but then switched over to a USB soundcard - there were two reasons.

1. The signal from the DAC always seemed quiet to me
2. Because my rig serves as a HTPC as well, accidently feeding a 5.1 Dolby signal into a DAC, with the volume up will cause a ear shattering rattle. Since I also run another cheaper set of 5.1s, forgetting to change the config was common - a USB soundcard doesnt seem to suffer from this.

As far as audio quality between the DAC and the USB, I can't tell the difference (except in max volume). Its proven to be perfectly reliable.

I use a Behringer UCA202. I know that a purist would say the jitter due to clocks, but I've never heard it.
 
its more than acceptable, in properly done, its superior to spdif.

we dont need to prove you wrong, all of the information is out there, plenty already on this site. the rapid uptake of it should tell you something and its not just convenience. its simply a method of shuffling the data. neither timing, nor tone is carried on the USB bus, only the data that represents tone and the timing is taken from a local clock, in the dac.

then add in the convenience and you have a winner
 
Last edited:
Using USB for audio data is fine. Using USB for audio sample timing is not. Unfortunately it seems that it can almost do it, which in a way is worse than if it totally failed because it leads some people to think that it can do it if only they find the right cable/software/DAC etc.

it doesnt use it for timing, it hasnt for years. there isnt a receiver on the market today that uses the timing except for stuff hanging over from many years ago, like the 270X series

this has been explained to you numerous times...

sorry for letting my frustration show through, the main reason for this frustration is that you are otherwise a pretty switched on guy from what I can tell, but I have personally explained this to you on at least 2 prior occasions and you keep on going back to it. posting what is now effectively misinformation in threads like this.

Using any element of USB for timing information is ancient history as audio technology goes, I have interfaces dating back nearly 4 years that are asynchronous and personally own 4 different USB->i2s convertors of varying quality, none of them use any timing info from USB.

it was the ones that did that gave USB audio a bad name and its still living it down through people like yourself, who had experience from back then and somehow cannot understand that this is not relevant anymore and hasnt been for quite a while.

you can pick up a quite decent convertor here on the forum, the Amanero combo384 card, which is async, with a USB clock and 2 dedicated audio clocks for 44.1x and 48x up to 384kHz PCM and DSD over USB. onboard power that can be regulated from USB, or supplied locally. you can get this for less than 50 euros for members here. there are several other async cards of varying prices and quality here on the forum.
 
Last edited:
it doesnt use it for timing, it hasnt for years. there isnt a receiver on the market today that uses the timing except for stuff hanging over from many years ago, like the 270X series

My estimate is vast majority of USB soundcards being manufactured and sold now is still adaptive. Async USB is easily available on the market today, but still a speciality for audiophiles. Windows even does not have a native (i.e. not third party) driver for USB audio v. 2 yet (which most of the new async DACs use).
 
windows probably never will or they would have already, but Thesycon and Centrance drivers are available to suit pretty much any device you buy. I think you'll find XMOS are more widespread than you think, since they offer such an extensive plug and play development platform. then you have Tenor TE7022L, CMEDIA CM6631A and a few other providers, including some FPGA and CLPD based custom devices that support async UAC2 or some bulk mode variants.

besides, since when do we judge the stock availability for drivers in an antiquated operating system? it has been native in mac since its inception and in linux for quite a while too. MS is being stubborn and IMO stupid/lazy, thats no reason to judge a technology as immature.
 
qusp said:
it was the ones that did that gave USB audio a bad name and its still living it down through people like yourself, who had experience from back then and somehow cannot understand that this is not relevant anymore and hasnt been for quite a while.
You tell me it doesn't happen.
phofman said:
My estimate is vast majority of USB soundcards being manufactured and sold now is still adaptive.
Others tell me it is still commonplace.

Who do I believe? I wasn't recommending doing it the wrong way; I was warning people to avoid it. Advising people not to do something because it doesn't work properly is hardly "misinformation" if it is simply impossible to do (because the parts are not, and never have been, available) and is certainly not "misinformation" if it is possible because it has been done and is still done by some people even though it doesn't work properly.

There are a lot of people out there who seem to believe that timing can be obtained from USB, and they play with their cables accordingly. I repeat: audio sample timing should not be obtained from USB. That is not misinformation, it is good advice. Even following the USB Audio spec will give serious timing errors, as that is only intended for low-fi applications like movie sound. Use USB for audio data only, with a big FIFO, then all will be well.

I don't know why you get so excited about this. I thought I was agreeing with you!
 
phoman is also mistaken, only lazy tacked on USB inputs are using any of that old technology commercially and here, where the question was asked about the current state of play, nobody would buy something like that and nobody would design using that old technology unless low budget was the dominant driving factor. its simply not a real problem with vaguely recent technology.

people will believe USB cables make a huge difference regardless of the USB method, much to my dismay, people still seem to feel it has an effect when there is a fifo and galvanic isolation in place ...
 
phoman is also mistaken, only lazy tacked on USB inputs are using any of that old technology commercially and here, where the question was asked about the current state of play, nobody would buy something like that and nobody would design using that old technology unless low budget was the dominant driving factor. its simply not a real problem with vaguely recent technology.

Let me repeat the OP point:

What a shock when I saw USB based sound cards and DACS

The OP talked about USB cards and DACs, not about speciality DACs for audiophiles.

Now, a user goes to a PC shop and asks for a USB sound card. Will he bring home async or adaptive? Well, the seller will have no idea what that means anyway...

A user needs a multichannel dac for his HT setup from notebook. What are his chances of buying async multichannel audio device?

If someone is not looking specifically for async USB audio device, he will most likely end up with adaptive one. That is reality, still today. Async is not default, and it takes a bit of effort for the buyer to avoid adaptive.
 
Last edited:
many decent semi-pro, or even hobbiest USB interface now, will be async, or custom bulk mode too. he also mentioned a 5 year absence and asked what had changed to make USB audio palatable. 5 years ago pcm270x receivers and the like, were pretty much ubiquitous and had quickly discounted USB as a serious audio interface for audiophile and pro-audio, for anyone not making custom hardware and drivers anyway. Firewire, or PCI was preferred in pro-audio, mostly for reasons of latency and bandwidth

I took the question as read, asked by an old DIYA member, diyer and audiophile, not a noob, to be related to quality audio gear, not plain vanilla tacked on budget USB crap, or otherwise quality hardware made by designers who dont know better, or dont put the effort into putting a proper USB input, instead taking the cheap and easy way out to round out the feature/buzzword checklist.

He was asking what had changed to make it successful. Anything not doing async, or custom bulk mode (both with clocking local to the dac) nowdays, in this market, is not taken very seriously and would not sell to anyone who has recent experience, or has taken the time to ask questions and async USB is such a buzzword all throughout the industry that it wouldnt take all that much to find one.

Such older legacy USB devices are widely known to be crap, even by new members here and audio hobbiests in general who have bought a dac in the last couple of years. Even if you ask among the teenagers over at head-fi, they will not buy a PCM270X based dac, they may not all know exactly why, but the spread of information and to some extent fashion, they know its inferior and there are multitudes of products for sale among the pages using async; from entry level through hi-end and portable, or home source gear.

how long do we consider that old technology be considered in questions like this? until the last cheap ebay crap, or utilitarian audio hardware design sells? how long must dated and nolonger relevant experience be passed on by experienced members who should know better, to those who may not know enough to search further after being warned of the evils of USB audio jitter? how long will such opinion/bias be spouted in new threads, in effect sustaining the ridiculous notion that spdif is superior?

In a properly designed dac, there should be little, or no difference on the output, but USB is certainly more convenient and flexible, no other interface has such direct and universal connection for PC audio and no other commonly available interface copes with stereo to multichannel 16-32bit 44.1-384+kHz PCM, as well as DSD
 
Last edited:
qusp said:
how long must dated and nolonger relevant experience be passed on by experienced members who should know better
Forgive me if I am mistaken, but that may be a reference to my comments.

Let me make one thing clear: I have absolutely no experience whatsoever of using USB for audio! I never claimed any experience. All I did was used my IT knowledge and looked at some of the USB specs. My conclusions were purely theoretical, but it seems that they agree with your findings. I continue to be puzzled why you seem so angry when someone agrees with you, and seem to want to silence a warning about how not to do something - while admitting that this method is still in use even though it should not be.

Are you a manufacturer in this field? If so, that might explain your apparent frustration with everyone else.
 
not you specifically mate, that was more a general comment, you dont hold the patent on this behavior ;) but yes I have seen you make comments to this effect, that point to timing and USB data clock jitter being a concern, when people ask about USB audio on this site.

there is years of this sort of info on the internet, by people that have some old experience that put them off and havent tried anything since, or alternatively have read only second hand psuedo-technical scaremongering jibberish.

People who have traditionally avoided USB audio for this reason, but are otherwise knowledgeable diyers, as well as diyers who are making their first foray into digital audio; come here and ask questions about USB audio and expect long term users of this site to give them a no ******** response based on relevant, recent first hand technical experience. Unfortunately often the same old information makes it into the threads and sustains it for another round. Just the mention of it, to someone who has already been reluctant and perhaps somewhat neurotic about taking this step into the 'new world' of digital audio, will easily sway them against it. Given it hasnt been a concern for years with the right products and now almost universally for modules here, thats unfortunate.

here on this site, there are many options and no interface that uses timing information directly is taken seriously, you would have to search harder to find a module promoted here that still uses these legacy type receivers.

I havent even mentioned the fact that a number of new DACs you might build to go with such a receiver, dont even use the master clock from the USB interface anyway.

I recommend you grab yourself an amanero to check it out for yourself, its cheap for DIYA members. here is the googlegroups list 48EU shipped is less than half retail price. ahh it'll be a bit more for you as you need to pay VAT. 56.46EU shipped
 
Last edited:
In a properly designed dac, there should be little, or no difference on the output, but USB is certainly more convenient and flexible, no other interface has such direct and universal connection for PC audio and no other commonly available interface copes with stereo to multichannel 16-32bit 44.1-384+kHz PCM, as well as DSD

Its about bloody time !

Cheers guys, thanks for the news..
:)
 
I rarely agree 100% with qusp, but in this case I have no clue what DF96 is arguing. All competent current USB audio being discussed on diyaudio at this time offer great flexibility and avoid the problems associated with USB audio eons ago.

To DF96: I am not a manufacturer in this field and do not intend to be, but I do believe that I understand the issues.
 
Last edited:
Audiophiles generally agree that PC audio works best when the operating system is stripped of all non-essential services, the network is disconnected, and the entire audio file is read into memory before playback begins. How does that compare with a USB DAC? The USB DAC includes a microprocessor that does real-time management of two USB endpoints: one controls the network protocol and the other formats and buffers the sample data. The 12 MHz oscillator for the USB network clock is always running and it’s asynchronous to the sample clock. There is rarely a moment when the processor is not attending to interrupts from the USB interface. With an I2S interface between the USB section and the DAC chip, the situation is bad enough. With S/PDIF between the USB section and a DAC chip in an external box, as is often done, you have the worst of both the USB and S/PDIF worlds.

A better approach is an S/PDIF connected DAC that derives all clocks from a local oscillator and exports a word clock to sync the source. All the clocks running in the DAC, including the S/PDIF PLL, are multiples of the sample clock.
 
hahaha, sorry, it sounds like you read too much fiction. there is just enough real info in the content to make it believable to some.

you tend to notice when a buffer underuns or overflows, its not a subtle thing. the above holds if you are getting dropouts, or segments repeating; then you might look at addressing what your resources are doing, if not, dont worry about it
 
Last edited: